ALMA-13B-Pretrain:可用于构建先进的语言翻译应用，项目采用两步微调范式，先基于单语数据微调，再用高质量平行数据优化，有效提升大语言模型的翻译性能。【此简介由AI生成】

ALMA（Advanced Language Model-based trAnslator，基于高级语言模型的翻译器）是一款基于LLM的翻译模型，它采用了一种全新的翻译模型范式：首先在单语数据上进行微调，然后使用高质量的平行数据进一步优化。这种两步微调流程确保了出色的翻译性能。

更多详情请参见我们的论文。

@misc{xu2023paradigm,
      title={A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models}, 
      author={Haoran Xu and Young Jin Kim and Amr Sharaf and Hany Hassan Awadalla},
      year={2023},
      eprint={2309.11674},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

@misc{xu2024contrastive,
      title={Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation}, 
      author={Haoran Xu and Amr Sharaf and Yunmo Chen and Weiting Tan and Lingfeng Shen and Benjamin Van Durme and Kenton Murray and Young Jin Kim},
      year={2024},
      eprint={2401.08417},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

from openmind import AutoTokenizer, AutoModelForCausalLM
import openmind
import torch
import torch_npu 
import argparse

def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument(
        "--model_name_or_path",
        type=str,
        help="Path to model",
        default="LF_AICC/ALMA-13B-Pretrain",
    )
    args = parser.parse_args()
    return args

args = parse_args()
model = args.model_name_or_path

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = openmind.pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)
sequences = pipeline(
   "<|im_start|>user\nDoes P=NP?<|im_end|>\n<|im_start|>assistant\n",
    max_length=256,
    do_sample=True,
    top_k=10,
    num_return_sequences=1,
    eos_token_id=tokenizer.eos_token_id,
)
for seq in sequences:
    print(f"Result: {seq['generated_text']}")

@misc{xu2023paradigm, title={A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models}, author={Haoran Xu and Young Jin Kim and Amr Sharaf and Hany Hassan Awadalla}, year={2023}, eprint={2309.11674}, archivePrefix={arXiv}, primaryClass={cs.CL} }

@misc{xu2024contrastive, title={Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation}, author={Haoran Xu and Amr Sharaf and Yunmo Chen and Weiting Tan and Lingfeng Shen and Benjamin Van Durme and Kenton Murray and Young Jin Kim}, year={2024}, eprint={2401.08417}, archivePrefix={arXiv}, primaryClass={cs.CL} }

from openmind import AutoTokenizer, AutoModelForCausalLM import openmind import torch import torch_npu import argparse def parse_args(): parser = argparse.ArgumentParser() parser.add_argument( "--model_name_or_path", type=str, help="Path to model", default="LF_AICC/ALMA-13B-Pretrain", ) args = parser.parse_args() return args args = parse_args() model = args.model_name_or_path tokenizer = AutoTokenizer.from_pretrained(model) pipeline = openmind.pipeline( "text-generation", model=model, tokenizer=tokenizer, torch_dtype=torch.bfloat16, device_map="auto", ) sequences = pipeline( "<|im_start|>user\nDoes P=NP?<|im_end|>\n<|im_start|>assistant\n", max_length=256, do_sample=True, top_k=10, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id, ) for seq in sequences: print(f"Result: {seq['generated_text']}")